Purpose: Vision is routinely used to walk from one nearby location to another, e.g. crossing at intersections. Motivated by the mobility problems of visually impaired people, we asked: 1) How accurately can people walk a straight line without vision? 2) Is this capacity affected by long-term vision loss? 3) Is such ability related to sensitivity in detecting curved paths? Methods: Five blindfolded sighted subjects and 5 blind subjects repeatedly attempted to walk along a straight 30'(foot) path. Deviations from the straight line were scored on a floor grid. In a 2nd experiment, the same subjects were guided along circular arcs of radii ranging from 30' to 120', and judged the direction of curvature (left or right). Threshold was defined as the curvature yielding 90% correct. Results: There were no clear differences between the blind and sighted groups. The mean absolute veer of straight-line walks calculated at 30' ranged from 1.4 to 5.7' across subjects. Analysis of the trajectories showed that in many trials, there were statistically significant linear and quadratic contributions to the veer. But these separate contributions were uncorrelated in both magnitude and direction (left vs. right). Mean threshold in the curvature discrimination task was a radius of ∼ 60'. We found no significant correlation between thresholds and any measure of veer in the straight-line walking task. Conclusions: Extensive experience with visual impairment does not influence the ability to walk a straight line or discriminate left from right- curving paths. The decoupling of linear and quadratic components in straight-line walking excludes simple random walk models for veering error. Surprisingly, accuracy in producing a straight-line walking path without vision is not related to perceiving the direction of curvature of a walking path.